Optimal synthesis and design of catalytic distillation columns: A rate-based modeling approach. (15th February 2021)

• Record Type: Journal Article
• Authors: Liñán, David A.; Bernal, David E.; Gómez, Jorge M.; Ricardez-Sandoval, Luis A.
• Is Part Of: Chemical engineering science. Volume 231(2021)
• Publication Date: 2021-02-15
• Subjects: Catalytic distillation -- Rate-based model -- Discrete-steepest descent algorithm -- Multiscale modeling -- Process intensification; Chemical engineering -- Periodicals; Génie chimique -- Périodiques; Chemical engineering; Periodicals; Electronic journals; 660
• DOI: 10.1016/j.ces.2020.116294
• Languages: English
• ISSNs: 0009-2509

Abstract

Highlights: An MINLP rate-based model for catalytic distillation (CD) design is presented. The rate-based model explicitly considers multiscale and multiphase phenomena in CD. MINLP was efficiently solved using a new Discrete-Steepest Descent Algorithm (D-SDA). A case study using a rate-based CD model was intensified using D-SDA. Results show that rate-based CD models are essential for optimal design of CD units. Abstract: This work presents the optimal synthesis and design of a rigorous catalytic distillation (CD) column that explicitly considers the multiscale and multiphase nature of this intensification process. A rate-based model that couples micro and macroscale events taking place inside the CD column is explicitly considered. The direct solution of such intensive optimization problems is challenging due to nonlinearities introduced by heterogeneous reactions, transport phenomena, and interactions between discrete and continuous variables. Also, combinatorial complexity involved in the selection of multiple structural decisions complicates the problem, e.g., distribution of reactive stages along the column, location of feed stages, and total number of stages. A discrete-steepest descent-based optimization framework recently developed is used to address the optimal synthesis and design of rate-based CD columns. A case study involving the production of ethyl tert-butyl ether (ETBE) has been considered. The results show that the multiscale events occurring in this process intensification (PI) unit cannot be ignored since they produce process designs different from those obtained with an equilibrium-based CD model. The results also show that neglecting multiscale phenomena may result in infeasible CD designs. The outcomes gained through this study illustrate the critical need to systematically consider multiscale and multiphase events in CD columns for the optimal design of realistic, cost-effective, and attractive process intensification (PI) systems.